Comparison of the trends in overall community-level mf prevalence predicted by the model with the observed baseline and post-intervention infection data collected longitudinally during the mass treatment programme.

Model outputs and observed infection data are shown for all the five PNG study villages. The observed overall community prevalence collected at a yearly interval from each study site are depicted by empty red squares together with their corresponding 95% binomial credible intervals (bars). The baseline data (at the 0th year in 1993–94) were collected soon before the start of the first mass drug administration round, and the five post-intervention mf prevalence data were collected annually a year after each mass drug treatment from the first through the 5th MDA round (see Table 1). Note that there are 500 modelled data points (shown by empty cyan circles) from the 500 best fits.